AISI Cyber CTF: leaderboard

AISI's 71-task private CTF suite where autonomous agents solve web exploitation, cryptography, reverse engineering, forensics, and privilege-escalation challenges.

Metric: Success Rate (%). Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.

Top models

#ModelScoreOverall rank
1GPT-5.3 Codex75#69
2Claude Opus 4.672#63
3GPT-5.2 Codex72#95
4GPT-565#98
5Claude Opus 4.563#83
6O354#124
7Claude Sonnet 4.545#142
8Claude Opus 439#163
9O138#183
10Claude 3.7 Sonnet34#249
11Claude 3.5 Sonnet30#350
12GPT-4 Turbo18#404
13Claude 3 Opus16#571
14GPT-4o15#343
15GPT-3.5 Turbo6#865

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=aisi-cyber-ctf · How It Works · Data refreshed daily, snapshot 2026-10-20.